Organic photochemistry. 40. Photochemistry of bichromophoric

Nov 1, 1978 - Harry Morrison , Alan Miller. Journal of the American Chemical Society 1980 102 (1), 372-373. Abstract | PDF | PDF w/ Links. Article Opt...
2 downloads 0 Views 2MB Size
1352

Journal of the American Chemical Society

(21)J. L. Holmes, G. M. Weese, A. Blair, and J. K.Terlouw, Org. Mass Spectrom., 12, 424 (1977). (22)P. J. Derrick, A. M. Falick, and A. L. Burlingame, J. Chem. Soc., Faraday Trans. 1, 1503 (1975). (23)R. Stockbauer and M. G. Inghram, J. Chem. Phys.. 65,4081(1976). (24)C. Lifshitz and M. Shapiro, J. Chem. Phys., 46,4912 (1967). (25)J. L. Holmes and A. D. Osborne. Int. J. Mass Spectrom. /on Phys., 23, 189 (1977). (26)J. H. Beynon, D. F. Brothers, and R. G. Cooks, Anal. Chem., 46, 1299 (1974). (27)M. A. Haney and J. L. Franklin, “Recent Developments in Mass Spectrometry”, K. Ogata and T. Hayakawa, Ed., University Park Press, Baltimore, Md., 1971,p 909. (28)H. M. Rosenstock and C. E. Melton, J. Chem. Phys., 26, 314 (1957). (29)R. D. Smith and J. H. Futrell, Org. Mass Spectrom., 11, 309 (1976). (30)M. L. Gross and F. W. McLafferty, J. Am. Chem. Soc., 93, 1267 (1971). (31)M. L. Gross, J. Am. Chem. SOC.,94,3744(1972). (32)J. L. Holmes and J. K. Terlouw, Org. Mass Spectrom., 10,787 (1975). (33)L. W. Sieck, R. Gordon, Jr., and P. Ausloos, J. Am. Chem. Soc., 94,7157

(1972). (34)M. L. Vestal, “Fundamental Processes in Radiation Chemistry”, P. Ausloos, Ed., Wiley-Interscience. New York, N.Y.. 1968,p 67-69. (35)S. G. Lias, R. E. Rebbert, and P. Ausloos, J. Am. Chem. SOC.,92,6430

/ 100:23 / November 8. 1978

(1970). (36)J. Sunner and I. Szabo, Int. J. Mass Spectrom. /on Phys., 25, 241 (1977). (37)D. P. Stevenson and C. D. Wagner, J. Chem. Phys., 19, 1 1 (1951);J. R. Nesby, C. M. Drew, and A. S. Gordon, J. Phys. Chem., 59,988(1955);W. H. McFadden and A. L. Wahrhaftig, J. Am. Chem. SOC., 78, 1572 (1956). (38)N. Dinh-Nguyen, R . Ryhage, S. Stallberg-Stenhagen, and E. Stenhagen, Ark. Kemi. 18. 393 11961). (39)N. Dinh-Nauven. R. kvhaae, . - and S. Stallbera-Stenhaaen, Ark. Kemi, 15, 433 (1966).. (40)8. J. Millard and D. F. Shaw, J. Chem. SOC.6,644 (1966). (41)J. H. Suzuki and K. Maeda, Can. J. Chem., 5 5 , 3124 (1977). (42)R . Neeter and N. M. M. Nibbering, Org. Mass Spectrom., 7, 1091 11973) (43)J L. Holmes and J. K. Terlouw. Can. J. Chem., 53, 2076 (1975). (44) H. C. Brown and S. Krishnamurthy, J. Am. Chem. SOC., 95, 1669 (1973). (45)A. I. Vogel, “Practical Organic Chemistry”, Longmans, London, 1948,p 258. (46)(a) A. Murray, Ill, and D. L. Williams, “Organic Synthesis with Isotopes”, Interscience. New York, N.Y., 1958,p 1482;(b) ibid., p 645. (47)E. C. Ashby and J. J. Lin, Tetrahedron Lett., 4481 (1977). \

\

- - - I

- I

Photochemistry of Bichromophoric Molecules. Photochemistry and Photophysics of 2-Methylenebenznorbornene and Related Nonconjugated Aryl Olefins in Nonprotic Medial Frank Scully, Tom Nylund, Fred Palensky, and Harry Morrison* Contribution from the Department of Chemistry, Purdue Unicersity, West Lafayette, Indiana 47907. ReceiGed April 25, 1978

Abstract: Three nonconjugated aryl olefins, having well-defined but varied interorbital angles between the chromophores, have been studied: 2-methylenebenznorbornene (MBN), 2-ethylidene- I-methylindan (EMI), and 1 -ethylidene-8,9-benzospiro[3.5]nonane (ESO). The latter two are new compounds and their syntheses are detailed. In all cases, photoelectron, ultraviolet absorption, and fluorescence emission spectra have been measured. The solution-phase photochemistry of these molecules, using 254-nm irradiation, has been elaborated; M B N reacts primarily via a “diverted di-.rr-methane” reaction while EMI, ESO, and 2-isobutylidenebenznorbornene (IBN) undergo a facile olefin E/Z photoisomerization. Xenon perturbation methods have been used to gain insight into the relative involvement of singlet and triplet states in these reactions, and it appears that both states participate. Singlet interaction between the chromophores is a function of their geometry, being greatest in MBN (where appreciable charge transfer occurs in polar media), reduced in EMI, and minimal in ESO. Interaction at the triplet level is complete in all cases, there being no aryl phosphorescence observed for any of the substrates. An important feature of both MBN and EM1 is extensive nonradiative decay from the singlet and triplet states; this decay is caused by the presence of the olefin moieties, does not involve olefin isomerization, and may well be general for nonconjugated aryl olefins.

The systematic study of bi- and polychromophoric molecules has generated considerable i n t e r e ~ t ,representing ~.~ as it does the logical extension of the intensive surveys of monofunctional compounds which dominated photochemistry until recent years. Our own efforts in this area have proceeded a t two levels: (a) we have been examining ground- and excitedstate interactions between functionalities in conformationally mobile, acyclic systems2 and (b) we have been studying these interactions in relatively rigid carbon frameworks. Each of the above has advantages with regard to obtainable information. Flexible systems provide a time-dependent spectrum of interchromophoric relationships which give rise to “preexcitation” and “postexcitation” interactions, and in fact may be used to obtain rate constants for bond rotations in chain^.^ By contrast, rigid molecules permit one to study interactions as a function of well-defined interchromophoric distances and angular relationship^.^ Among the functional group pairs examined in these laboratories, the aryl olefins have proven particularly fruitful, giving rise in acyclic arrays to both triplet and singlet interactions upon excitation of the benzene ring. Thus 6-phenyl0002-7863/78/1500-7352$01 .OO/O

2-hexene provided the first suggestion of, and evidence for, aryl olefin singlet exciplexes as precursors to cycloadducts.6 Following excitation, the two functionalities achieve a collinear arrangement (allowed for by the trimethylene “bridge”), which permits rapid (4 X lo8 s-I) complexation, almost complete fluorescence quenching, and eventual photochemistry (eq I ). hu

Ph(CH,)&H=CHCH, (2)

By contrast 1-phenyl-2-butene exhibits normal singlet photophysics but undergoes a facile, predominantly triplet derived, E/Z photoisomerization (eq 2).7 hv

PhCH,CH=CHCH,

PhCHLCH=CHCH3

(2)

254 n m

0 1978 American Chemical Society

(E)

(2)

Morrison et al.

/

Photochemistry of 2-Methylenebenznorbornene

To further define the stereoelectronic requirements for aryl olefin excited- and ground-state interactions, we have synthesized a series of polycyclic molecules having different functional group orientations. These include 2-methylenebenznorbornene (MBN), 2-ethylidene- 1-methylindan (EMI),

MBN

EM1

ESO

and l-ethylidene-8,9-benzospiro[3.5]nonane(ESO). Our discussions will be restricted to the photochemistry and photophysics of these compounds in nonprotic media;s an elaboration of our observations in protic media9 is presented in the accompanying paper. l o

Results Photochemistry of 2-Methylenebenznorbornene (MBN). Irradiation of a 0.01 M solution of M B N in n-hexane with a low-pressure mercury lamp results in the formation of four primary photoproducts (1-4) isomeric with the starting material." The major product, l , is a new compound and has been identified as 1-methylene-3-vinylindan by spectroscopic analysis (cf. Experimental Section). Compound 2 (benzo [ 6.71 bicycle[ 3.2.11 octa-2,6-diene), 3 (3-allyl- 1H-indene), and 4 (1-allyl-1H-indene) have been previously characterized and were identified by comparison of spectra with authentic, independently synthesized samples. The reaction is summarized in eq 3.

n

H hv 254 n m

1

2

4

3

Photochemistry of 2-Methylene-d-benznorbornene. MBN, deuterated in the methylene position, was readily prepared using triphenylmethyl-d3-phosphonium bromide and benznorbornen-2-one. Photolysis was conducted under conditions identical with undeuterated material, products were isolated by VPC, and the location of the deuterium atoms was ascertained by N M R (see Experimental Section). The reaction is depicted in eq 4.

n

Quantum Efficiencies of Formation of Photoproduct13and Disappearance of MBN. Quantum efficiencies were determined in low-conversion runs and are as follows: 1,6.3 X 2, 1.5 X 3, 1.3 X 4, 0.8 X The separately determined quantum efficiency for disappearance of M B N is 1 3 x 10-3.

7353

Effect of trans-2-Heptene on the Photochemistry of MBN. T o confirm that the photochemistry caused by direct irradiation does not involve intermolecular aryl-olefin interactions, M B N was irradiated in the presence of an external olefin, frans-2-heptene. The use of molar quantities of the olefin 12-59-fold greater than M B N led to no observable quenching of the formation of 1. In a separate experiment, the solutions were analyzed for fortnation of cis-2-heptene. Using 5.4 X to M M B N and olefin concentrations from 6.8 X 19 X 10-3 M, prolonged irradiation times were required before measurable amounts of cis isomer could be detected (Le., under comparable photolysis conditions, of the order of 7% conversions to 1 could be achieved in -6 h, whereas the 6.8 m M heptene tube required -30 h for -9% conversions to cis-2heptene). Since secondary photoproduct formation would be significant with such long reaction times,' we interpret these results merely as confirmation of the short-lived, noninterceptable nature of the M B N triplet. These observations are substantiated by the lack of M B N phosphorescence (see below). Photosensitized Rearrangement of MBN. W e have demonstrated that the anti-Markownikoff addition of acetic acid to M B N proceeds from the excited singlet ~ t a t e . ~This . ' ~ reaction can then serve as a probe for singlet energy transfer from a prospective triplet sensitizer. Thus, hexane solutions of toluene and MBN, in which toluene absorbed essentially all 254-nm incident radiation, gave rise to M B N rearrangement, but control studies indicated that acetic acid addition to MBN occurred as well. By contrast, p-xylene did not sensitize addition of acetic acid but did effectively cause M B N rearrangement. The product ratio, for 70% loss of MBN, is 1,90%; 2,5%;3,4%; and 4,0%.These numbers may be compared with the ratio from low-conversion, direct irradiation quantum efficiency data: 1, 62%; 2, 19%; 3, 13%; 4, 5%. A quantitative comparison of these data is beclouded by the ambiguities of photosensitized secondary reactions (for example, there is a 9% formation of the di-T-methane photoproduct of 2,l which explains the diminished presence of 2 in the sensitized run). However, the M B N triplet is clearly capable of giving rise to 1,2, and 3, and the absence of 4 and the increased proportion of 1 seem statistically significant. Xenon Perturbation Studies on MBN. Varying amounts of xenon were added to degassed hexane solutions of MBN, and the yields of photoproducts compared with those obtained in the absence of xenon. The formation of all four photoproducts (1-4) was enhanced by xenon, the increases ranging from 50% (1) to over 100% (4). The degree of fluorescence quenching was measured simultaneously and plotted by the method of Fleming, Quina, and Hammond,14 i.e., [(Fo/F) - I ] vs. [(Fo/F)(P/Po) - I ] . In these expressions, FO = fluorescence without xenon, F = fluorescence with varying xenon concentrations, PO= product formation without xenon, and P = product formation with varying xenon concentrations. Least-squares plots of the data of each of the products are shown in Figure I ; the slopes (which equal 4isc when the product probe is 100% triplet derived; cf. the Discussion section) are 1, 0.52 f 04; 2,0.44 f 0.06; 3, 0.41 f 0.04; and 4, 0.32 f 0.03. Photochemistry of 2-Isobutylidenebenznorbornene (IBN). IBN was prepared by the Wittig olefination of benznorbornen-2-one, there being formed a 90:lO mixture of Z and E olefins. Larger quantities of the E isomer were prepared by 254-nm irradiation of the synthetic mixture. The Z and E isomers were separated by preparative VPC and characterized by I3C N M R (cf. Experimental Section). Irradiation of either isomer in hexane solutions using 254-nm light rapidly led to geometrical i s o m e r i ~ a t i o n (eq '~ 5). Quantum efficiencies for these reactions are ~ Z - E = 0.10 and $,+z = 0 .08 (under the conditions of these measure-

7354

Journal of the American Chemical Society

/

100:23

/ November 8, 1978

I .!

-

I. (

LALL

Figure 1. MBN xenon perturbation plots: 1 (A),2 (O), 3 (X), 4 (0)

Figure 2. Xenon perturbation determination of @isc for BN.

ments, formation of the other products is negligiblei5). The E/Z photostationary state was determinedI6 to be 1.1 1 and was unchanged by the presence of xenon.

1

Z.IBN

hu

6

E-IBN

Determination of hsrfor Benznorbornene (BN). This number was obtained by the xenon perturbation method of Carroll and Quina.” This involves simultaneously measuring the effect of xenon on a molecule’s fluorescence (quenched by xenon) and its triplet-sensitized isomerization of an external olefin (enhanced by xenon). The equation is as follows:i7

F’ = fluorescence without olefin or xenon, Fo = fluorescence without xenon, F = fluorescence with olefin and varying xenon concentrations, PO = olefin isomerization (cis-2-heptene) without xenon, and P = olefin isomerization with varying xenon concentrations. The data are plotted in Figure 2, with the least-squares-derived slope (&-) equal to 0.59. W e had earlier done a more routine @isc determination on 2-methylbenznorbornene (MeBN), using trans- 2-heptene as a triplet counter. Using a decay ratio equal to that for the toluene-sensitized isomerization of heptene, we obtained a 4iSc

Figure 3. IBN xenon perturbation plot.

of 0.62 f 0.04, in reasonable agreement with the benznorbornene value. Xenon Perturbation Studies on IBN. Using a procedure identical with that discussed for M B N (cf. Figure l ) , but employing Z to E photoisomerization of IBN as the product probe, gave the data plotted in Figure 3. The slope is 0.46. Spectroscopy of MBN. A. Absorption and Fluorescence. These spectra, taken in hexane a t room temperature, are presented in Figure 4, with data for MeBN by way of comparison. &values for these and related compounds are given in Table I together with singlet lifetimes. The reductions in +f for the unsaturated bicyclics at room temperature are observed as well as 77 K, Le., the Cpf ratio for MBN:MeBN = 0.38 and 2 6 = 0.50, at 77 K. B. Phosphorescence. None of the aryl olefins showed phosphorescence in an isopentane glass at 77 K. Both MeBN and 6 phosphoresce, with efficiencies of 1.70 and 0.7 1 relative to toluene. C. Photoelectron Spectroscopy. Vertical ionization energies for MBNI9 and related molecules are presented in Table 11. Preparation of 2-Ethylidene-1-methylindan(EMI). EM1 was prepared by the sequence outlined in Scheme I. This route provides a mixture of the isomers, with >98% E . Enrichment of the Z isomer was accomplished by irradiating the synthetic mixture. The isomers were separated by preparative VPC and structures assigned using I3CN M R spectroscopy (cf. Experimental Section). Photoisomerizationof EMI. Irradiation of either EM1 isomer led to rapid E/Z interconversion22 (eq 6). The quantum

E

-

254 nm

z

efficiency for E Z is 4E-z = 0.146 f 0.012 and the photostationary state, (E/Z)pss= 0.88 f 0.02. A combination of these data and the extinction coefficients a t 254 nm ( E , 401 M-l cm-I; Z , 409 M-’ cm-I) gives ~ Z - E = 0.13. The photostationary state changes under the influence of xenon, to become ( E / Z ) p s s= 0.99 f 0.02. Withp-xylene sensitization, ( E / Z ) p s=s 1.10 f 0.01. Photolysis of E-EM1 in the presence of an equimolar concentration of cis-2-heptene gave 4% conversion to Z-EM1 without any detectable formation of trans-2-heptene. The quantum efficiency for disappearance22 of EM1 = 0.01. Xenon Perturbation Studies on EMI. Use of the procedure outlined earlier for M B N and IBN, in this case monitoring E Z photoisomerization as a function of xenon concentration,

-

Morrison et al.

/ Photochemistry of 2-Methylenebenznorbornene

Table I. Quantum Efficiencies of Fluorescence and Singlet

7355

compd

-100

1200L

MBN -

Lifetimes of M B N and Related Molecules @fa

17,

nsb

0.24

13.5

0.08

4.5

1000

MsBN

-

______

- 80

CH,

11.0

X (nm) Figure 4. Absorption and fluorescence spectra for MBN and MeBN, in hexane at room temperature. (Note that the MBN fluorescence is scaled up by a factor of 2.92).

0.25

14.0

0.15

8.5

c

6

30

2

All &values were determined relative to toluene (@f= 0.14).18 These values are considered accurate to f l ns.

Figure 5. E-EM1 xenon perturbation plot.

Table 11. PES Values for MBN and Related Molecules

compd

NaBH,CN pH DMF 3.8, /sulfolane 110 OC

+& +& $@c \

\

E-EM1

CH3

2-EM1

gave the data plotted in Figure 5. The least-squares-derived slope is 0.61 f 0.01. Determination of hSc for 1-Methylindan.This measurement was done by the xenon perturbation method,” using cis-2heptene as described above for benznorbornene. The data are plotted in Figure 6, with the least-squares-derived slope, @isc = 0.59 f 0.002. Spectroscopy of EMI. A. Absorption and Fluorescence. These spectra, taken in hexane a t room temperature, are presented in Figure 7 together with data for 2-ethyl- l-methylindan (EMIH) as a model monochromophoric system. The @ i s for E - and Z-EM1 are both 0 . 2 2 , whereas E M I H has @f = 0.34. The lr’s reflect this difference, being 24 ns for E M I H and 18 ns each for the EM1 isomers. B. Phosphorescence. Neither of the EM1 isomers shows phosphorescence in an ethanol glass at 77 K, whereas E M I H phosphoresces with an efficiency 0.77 that of toluene. C. Photoelectron Spectroscopy. Vertical ionization energ i e ~ *of~ the structurally simpler EM1 analogue 2-methy-

@5

vertical ionization potential, eV

assignment

E

benzene S benzene A olefin

8.4220 8.93

benzene S benzene A

9.0421

olefin

9.0 1 2 1 9.38

olefin olefin

8.40

BK

l e n e i r ~ d a n ,and ~ ~ related molecules are presented in T a b l e 111. Preparation of l-Ethylidene-8,9-benzospiro[3.5]nonane (ESO).ESO was prepared by the sequence outlined in Scheme 11. This route provided a mixture of isomers, with a Z:E ratio of 95:5. The isomers were separated by preparative VPC and assigned structures using IH NMR (cf. Experimental Section). Larger quantities of the E isomer could be prepared by irradiation of the synthetic mixture.

7356

Journal of the American Chemical Society

Ot 20 00

30-

I

/

100:23

/

Nocember 8, 1978

-130

Absorption

clLol!J-

-20-

L ILL0 \

300

40

I O

20

40

30

50

Figure 6. Xenon perturbation determination of $isc for 1-methylindan.

310

Scheme I1

Table 111. PES Values for 2-Methyleneindan and Related Molecules

1. Me,SO/KOHb

vertical ionization potential, eV

cornpd D

C

83

H

+

assignment

8.61 9.14 9.45 8.46*O 9.04

benzene S benzene A olefin benzene S benzene A

9. I5?‘

olefin

Photoisomerization of ESO. Irradiation of either E S O isomer led to rapid E/Z interconversion (eq 7). Quantum ef-

XO

X (nm) Figure 7. Absorption and fluorescence spectra for EM1 and EMIH, in hexane at room temperature. (Note that the E-EM1 fluorescence is scaled up by a factor of 1.53).

Ph’t-

Z.HBF,

BF,-

Ph,P=CHCH,,

3 E-ESO

8 +

HJc& \ 2-ESO

glass. Rather surprisingly, the monochromophoric analogue, M S O H , also shows no phosphorescence. C. Photoelectron Spectroscopy. Vertical ionization energ i e of~ MSO, ~ ~ MSOH, and related molecules are presented in Table IV.

Discussion A. Photochemistry of MBN. 1. Products and Mechanism. The photochemical rearrangements of M B N (eq 3) are the E Z most deep-seated of the transformations detailed in the Results section; photolysis of EM1 and ESO leads primarily*’ to E ficiencies are -.z = 0.200 f 0.004 and &-E = 0.204 f 2 interconversion (eq 6 and 7). The experiment with deute0.006. The photostationary state, ( E / Z ) p s = s 1.00 f 0.01. rium-labeled MBN (eq 4) allows for a rationalization of the Extended irradiations of equimolar concentrations of 1photolysis in terms of a “diverted di-x-methane” reactionsa met hylene-8,9-benzospiro [ 3.51 nonane (MSO) and cis-2(Scheme 111). Excitation is followed by the di-aheptene showed no measurable heptene isomerization. methane bond-making and bond-breaking steps to give the Xenon Perturbation Studies on ESO. Xenon had essentially 1,3-diradical8. The ultimate formation of 2,3, and 4 requires no effect on either ESO Z --* E or E Z isomerization, althat the “normal”, but highly strained, di-9-methane product though xenon did quench ESO fluorescence. The model monochromophoric compound, 5,6-benzospiro[3.5]nonane, 9 is formed as a transient.26 However, the major pathway is fragmentation of 8 to the carbene 10, which rearranges to 1. was prepared by the catecholborane reduction and decompoSimilar diversions from di-x-methane cyclization have been sition of the tosylhydrazone of 8,9-benzospiro[3.5]nonan1observed for dibenznorbornadiene~~’ and triptycenes.28 one. Attempts to measure &c for this compound by xenonWhen an appropriate label is present, such as the isopropyl enhanced sensitization of a 2-heptene isomerization (as degroup in IBN. geometrical isomerism of the olefin becomes scribed above) were unsuccessful; again fluorescence evident (eq 5 ) . Isomerization is far more efficient than rearquenching by xenon was observed, but no effect on 2-heptene rangement, for of MBN is 0.013 where ~ E - Z ~ Z - E of isomerization was noted. IBN totals 0.18. Spectroscopy of ESO. A. Absorption and Fluorescence. The 2. Excited State Multiplicity. In principle, both rearrangeabsorption (>220 nm) and fluorescence spectra of E - and ment and olefin isomerization could arise from singlet and/or Z - E S O are essentially identical in band shape and intensity; triplet precursors. Formation of 1-4 primarily from the singlet they are also virtually superimposable on those of a monochromophoric model, l-methyl-8,9-benzospiro[3.5]nonane is to be expected, since this is observed for di-n-methane rearrangements involving an isomerizable olefin.25 Conversely, ( M S O H ) . The c$f for the methylene analogue, MSO, was olefin isomerization is not normally competitive with singlet measured using toluene as a standard and found to be 0.18. The di-a-methane ~ h e m i s t r y , *and ~ . ~one ~ would therefore anticisinglet lifetimes of ESO, MSO, and the related monochropate that the IBN isomerization is primarily due to a triplet, mophoric models are all 30.5 ns. either via reversibility in the bridging step (7)25or through B. Phosphorescence. As with the other aryl olefins discussed decay to a triplet localized a t the double bond (intramolecular above. ESO shows no phosphorescence a t 77 K in an ethanol 254 nm

*

-

+

Morrison et al.

7357

/ Photochemistry of 2-Methylenebenznorbornene Table IV. PES Values for MSO and Related Molecules

Scheme I11

cornpd I

ED,

CD,. 7

8

68 >IS0

vertical ionization potential, eV

assign mc n t

8.40

benzene S benzene A olefin

;:E

benzene S benzene A >ISOH

8.96

benzene S benzene A

9.1 921

olefin

8.4420

$BYisc l T M B N 4MBNlsc

=

I TBN

= 0.20

The value so determined38 is 4iSc= 0.20; since xenon enhances the M B N rearrangement, the data fall in the inconclusive category.35 An alternative approach is to quantitate the xenon enchancement of product formation and the xenon quenching of fluorescence as a function of relative xenon concentration. This method has been successfully employed by Hammond and co-workersI4 using bromocyclopropane as the perturbing agent. When the probe product is 100% triplet derived, the equation

triplet energy transfer).’ The assignment of multiplicity to excited state precursors in bichromophoric aryl olefins is complicated by the very short triplet lifetimes (none of the unsaturated molecules under discussion phosphoresces). The lack of observable quenching by the usual triplet quenchers is [ ( F o / F )- 1 I = @isc[(Fo/F)(P/Po) - 11 ( 1 1) therefore inconclusive. W e have turned to xenon perturbation applies (and may in fact be used to obtain However, it and photosensitization experiments for an insight into this is easily demonstratedlb that when the probe product is only problem. partially triplet derived, a modified equation is applicable: As regards sensitization, p-xylene has been used successfully to selectively create the MBN triplet state.30The lack of singlet [ ( F ~ / F-) 11 = 4isc [ ( ~ o / ~ ) ( p/ ~11o ) ( 1 2) energy transfer is demonstrated by xylene’s inability to induce 4 P a known M B N singlet reaction (anti-Markownikoff addition where dP is the total quantum efficiency of probe product of acetic acid to the double bond).sc~9,10,31 That xylene sensiformation and 34pis the efficiency via the triplet manifold (i.e., tization does yield 1,2, and 3 therefore assures that an MBN 4isc&~).Thus the slopes of such plots, for a multiple product triplet is a t least capable of leading to these pr0ducts.3~ reaction, must increase as the singlet component increases; in However, one must ultimately depend on perturbation experiments for information concerning the direct p h o t ~ l y s i s . ~ ~no case can the slope be less than 4isc.Plots for MBN are given in Figure 1 (Results section) with the slopes ranging from 0.32 Such experiments are not without potential ambiguity, for to 0.52. If one again assumes 4iSc= 0.20, all products are when both singlet and triplet intermediates participate in a partially singlet derived (cf. Table V). The same conclusion reaction, whether there is heavy-atom quenching or enis required f o r 1BN geonietrical isomerization. 40,41 The last hancement depends on the relative quantum efficiencies of observation is particularly noteworthy, and could be explained singlet reactivity ( ~ R s ) , intersystem crossing (@isc), and inby invoking reversibility in a singlet-derived bridging step (7 trinsic triplet reactivity ( 4 R T ) . 3 4 A kinetic a n a l y ~ i s 3indicates ~ in Scheme III).40b that (1) when 4isc1 0.05, enhancement of product formation 3. Nonradiative Decay of the MBN Excited States. It is in(by, for example, xenon) is unambiguous and requires that the product be predominantly triplet derived (Le., #Ji&RT > 4 ~ s ) ; structive to calculate the total quantum efficiency for MBN product formation out of the singlet manifold, using the per(2) when 4isc I0.05, quenching unambiguously confirms a centages given in Table V and including olefin isomerization ~). predominantly singlet derived product ( ~ R S> 4 i s c 4 ~Other as indicated by the IBN data. To the sum (0.10) may be added combinations of @isc and enhancement or quenching are in$f (0.08) and the assumed &c (0.20). Clearly only a s m a l l conclusive in the absence of +RS and +RT data.35The 4isc value portion (38%) of M B N singlets is accounted for, the nonfor M B N is obviously critical, but because of the very short radiative decay via internal conversion being l$d = 0.62. By triplet lifetime, must be obtained indirectly. W e have found contrast, l 4 d for dihydro models equal ca. 0.17 (1 - (4f 4isc for a monochromophoric analogue, benznorbornene (BN), disc)).In terms of rate constants, ‘kd for MBN must be 1.4 X to be 0.59. If one assumes that the kist of BN is unchanged IO8 s-I, some tenfold greater than Ikd (1.3 X lo7 s-I) for upon introduction of the double bond,36 4isccan be calculated MeBN.42 Introduction of a double bond at the 2 position has by brought about a new singlet decay niode which (a) presumably 4BNisc = kBUisclTBl\i involves benzo-vinyl interaction; (b) does not involve rotation (8)

$

+

7358

Journal of the American Chemical Society

/

100:23

1 November 8 , 1978

Table V. Singlet Contribution to MBN Rearrangement and IBN Isomerization from Xenon Perturbation Studies

a

photoproduct

slope

%singleta

1

2

0.52 0.44

3 4 IBNZ4E

0.41 0.32 0.46

62 55 51 38 51

Assuming & = 0.20.

about the double bond; (c) undoubtedly involves significant charge t r a n ~ f e r ;(d) ~ ~does ~ ~not ~ ~involve . ~ ~ the formation of a full bond between the 2 and 6 carbons.I0 We would represent the form of this interaction in structural terms as 11. Obviously,

Figure 8. Orbital splitting diagram for MBN using corrected basis levels (see text).

With regards to the MBN ground state, the PES data have been presented in Table I1 (Results), with the S and A as-

11

a t some stage full 2,6 bonding does develop and brings about single-derived olefin isomerization and di-.rr-methane chemistry. Interception of an aryl excited singlet state by a nearby olefin, resulting in nonradiative decay without appreciable olefin isomerization, is well p r e ~ e d e n t e d . ~ Perhaps ~ ’ ? ~ ~ more striking then is the evidence f o r a similar situation at the triplet level. The modified form of the xenon perturbation equation (eq 12) can be used to provide $RT, the intrinsic triplet reactivity:

signments referring to a symmetry plane bisecting the benznorbornenyl ring system. Our treatment follows that of H e i l b r ~ n n e r in , ~ which ~ the basis energies are values from appropriate models, modified to reflect the inductive (through-bond) effect of the functional groups on one another. The inductive correction for the MBN models is the difference

+

Using the d,, data for products 1-4 $ E * z , and the slopes in Table V, one calculates a total $RT of 0.41; Le., despite the total absence of phosphorescence, only 41% of the M B N triplets are accounted f o r . 44 Since nucleophilic addition to MBN is confined to the singlet ~ t a t e , ~ the ~ , most ~ , ~ eco, ~ ~ nomical rationale is to invoke a triplet equivalent p structure 11 which has a negligible charge-transfer component. It must be emphasized that this interaction is competitive with localization of triplet energy at the double bond45 (with the latter viewed either as “triplet energy transfer” or T2 TI decay), since localized (acyclic) olefinic triplets decay entirely by rotation. We are forced to similar conclusions for the indan, EMI. A summary of MBN photochemistry and photophysics is given in Scheme IV. 4. Interaction between the Chromophores. To what extent is there interaction between the olefin and aromatic moieties in the ground state? What is the effect of the double bond on the SO SI,transition and on SI itself? Despite the extensive research on bichromophores, these questions remain difficult to answer. We have utilized photoelectron, ultraviolet absorption, fluorescence, and phosphorescence spectroscopy to address these problems.

-

-

MBN

hv c

‘(MBN)*

20%

3(MBN)* 5 9 y

MBN -k

hv

MBN 1, 2, 3, 4,E*Z

eV

MBN

I“’”

1, 2, 3, 4 Ec’Z

MBS

BX

8.40 S 8.90 A 9.23 7-r

8.42 S 8.93 A

12

9.04 7-r

between the average of the energies of BN erage of the energies of MBN: (8.42

+ 12 and the av-

+ 8393 + 9.04) - (8.40 + 8390 + 9.23

)

= -0.05 eV

The correction (-0.05 eV), applied to BN and 12, yields basis energies of 8.47 (S), 8.98 (A), and 9.09 (9) eV. Comparison of these data with those of M B N shows that the P level has been split by 0.14 eV, in good agreement with the sum of the benzene S and A splittings of 0.15 eV. An orbital interaction of 0.29 eV is not inconsequential, and approximates that calculated in like fashion for 13.21The interaction energies for the endocyclic analogues 1447and 1548are expectedly higher,

MBN orbital splitting, eV

Scheme N

A

S

0.29

13 0.30

14 0.46

15 0.86

but it is somewhat surprising that the decrease from 15 to 14 is not reproduced in the exocyclic series. Despite the calculated interaction energy, the HOMOS of MBN and BN have almost identical energies, with the major perturbation occurring a t the M B N olefin MO.

Morrison et al.

1 Photochemistry of 2-Methylenebenznorbornene

7359

Table VI. Aryl Olefin Interorbital Angles and Distances and The U V absorption spectra of MBN and MeBN are preSinglet Lifetimes sented in Figure 4 (Results). There is a small (-2 nm) bathochromic shift of the Blu band and some diminution of distance, % ’ 7 Of intensity upon introduction of the double bond. The spectrum compd angle,degU A‘ I T , ns satd model of MBN has been discussed in detail,49with the conclusion that I major perturbations are restricted to the Elua and Bl, transi180 2.2 18.0 15 t i o n ~Similar . ~ ~ conclusions have been reached using C D and O R D methods.53 Nevertheless, there is general a g r e e m e ~ ~ t ~ ~ . ~ ~ EM1 that the B z transition ~ (which is responsible for the photochemistry under discussion) is slightly perturbed. The fluorescence spectrum (Figure 4) of MBN is much more dramatically altered. Both &and I 7 are reduced by 66% relative to MeBN (Table I); since $r = k f l 7 , kf is essentially unchanged (as one would anticipate from the absorption data). 125 2.0 11.0 75 For reasons outlined above, the diminution in &and I T can be ascribed to a much increased nonradiative decay mode which does not involve olefin rotation. It is clear from Figure 4 that the fluorescence of M B N is, like its absorption spectrum, 90 3.0 30.0 100 slightly red shifted from that of MeBN. In sum (1) the HOMO and LUMO energies of MBN aptMSO1 proximate those of model systems; ( 2 ) the SOQ SItransitions 0 Defined in the text show a small butfinite coupling o f t h e chromophores; ( 3 ) the SI state is characterized by a markedly increased rate of nonradiatice decay. The ultimate question is whether 11 repScheme V resents SI, or is itself derived from a more “benzene-like’’ hv 44% initially formed singlet state. We believe that latter possibility EM1 to be realistic, even with the relatively fixed nuclei of a bicyclic system; a more complete discussion may be found in the accompanying p a ~ e r . ’Its ~ ,properties ~~ place MBN somewhere EM1 EM1 EGZ EM1 E*Z between 6-phenyl-2-hexene (where there is exclusively a post+ excitation interaction6) and alkyl phenylacetates (where there hv is gross spectral distortion and the possibility of delocalized excitation into a “ s u p e r c h r ~ m o p h o r e ” ~ ~ ) . B. Photochemistry and Photophysics of EMI. A primary 41scyields a l$d for E M I H of 0.07. For EMI, & (0.22), $lsc objective of this study was to examine the photochemistry and (0.44), and I $ E = z (0.08) leave a l$d of 0.26. Introduction of photophysics of aryl olefin bichromophores as a function of the double bond leads to an additional 19% nonradiative decay, geometry. The “interorbital angle” (i.e., the angle between the or in terms of ’ k d , 1.4 X IO7 s-l (EMI) vs. 2.9 X IO6 s - I nodal planes of the two P systems, a t a conformation inter(EMIH). The fivefold increase in Ikd may be compared with mediate between the conformational minima) and the “inthe tenfold increment for MBN. Though EMl’s geometry is terorbital distance” (Le., the closest distance between a carbon clearly not as well suited for the partial bonding depicted in 11 atom of the olefin, and an aryl carbon or aryl C-C bond, again (above), there is evidence for bond formation across the fivein an intermediate conformation) are given in Table VI for membered ring from the photoinduced skeletal rearrangement several olefins of interest. As one would expect, photophysical of indene^.^^.^^ Once more, the triplet mimics the singlet; there properties (exemplified by I7 in Table VI) vary appreciably; is no phosphorescence and 3 $ ~ s = z 0.20. Thus 55% of the a discussion of the photochemistry of EM1 and MSO follows EM1 triplets are unaccounted for and decay is competitive with (see Results for the synthesis of these compounds). triplet localization a t the double bond. Our observations are By contrast with MBN, photoisomerization is the primary’? summarized in Scheme V. photoreaction of EM1 (eq 6). The quantum efficiencies ($ELZ Treatment of the PES data (Table 111) as for MBN yields = 0.15, $z--E= 0.13) are comparable to those observed for a -0.23 eV through-bond inductive effect, which applied to IBN and the parent acyclic molecule, I-phenyl-2-butene. As the model systems gives basis energies of 8.69 (S), 9.27 (A), with MBN, there is evidence for singlet involvement in the and 9.38 (P)eV. For 2-methyleneindan, the corresponding isomerization. For example, the photostationary state (E/Z values are 8.61, 9.14, and 9.45 eV, respectively; symmetry = 0.88) is changed by xenon ( E / Z = 0.99) in the direction of permits only a benzene (s)-P interaction which totals 0.15 eV. that observed for xylene triplet sensitization (E/Z = 1.10).56160 Though this number is relatively small, the H O M O of the aryl The extent of singlet participation can again be estimated from olefin is stabilized by 0.1 5 eV by comparison with indan; MBN xenon perturbation studies, using the slopes of Figures 5 (0.61) has a greater calculated orbital interaction but virtually unand 6 (0.59), and the $,sc (0.44) derived from IT’S for EM1 ( I 8 changed H O M O . ns) and 1-methylindan (1-MI) 24 ns, from EMIH). See eq As with MBN, there is a small (ca. 2 nm) red shift in the UV 16-1 8.61 absorption spectrum of EM1 (Figure 7). Though the fluorescence intensity change is less marked ( E M I / E M I H = 0.64; MBN/MeBN = 0.33), the band shape is more distorted and red shifted. The decrease in qbf is attributable to ‘ k d ,kf being (4p/3$p)= 0.61/0.44 = 1.39 virtually unchanged. As noted above, EM1 gives no detectable (17) phosphorescence. % singlet = 1 - ( ) I $ ~ / $ ~=) 28% (18) In sum, much of what has been observed for the photoisomerization and photophysics of MBN is reproduced with The 28% so calculated may be compared to the 57% calculated EMI. The degree of interaction between the chromophores has, for IBN, the reduced singlet participation for EM1 quite poshowever, been reduced by the collinear (EMI), as opposed to sibly reflecting the higher relative I T of this molecule (Table angular (MBN), relationship in the two molecules. VI). As regards the singlet state, using 0.34 and 0.59 for $f and

6

& 68

7360

Journal of the American Chemical Society

\

\

Figure 9. Orbital splitting diagram for 2-methyleneindan using corrected basis levels (see text).

C. Photochemistry and Photophysics of ESO. Of the cyclic substrates discussed herein, the spiro-fused ESO molecule is unique in that its I T is identical with that of the model system (Table VI). Nevertheless, ESO does undergo facile geometrical photoisomerization (eq 7); in fact, the quantum efficiencies ( @ L + z = @ Z - E = 0.20)64are the highest observed in this series. Evidence concerning reaction multiplicity is sparce. Though xenon quenches the fluorescence of both ESO and monochromophoric models, there is inexplicably no observable change in efficiency of internal or external probe isomerization. One can say that the I T data, which imply minimal singlet interaction, are supported by the absorption and fluorescence spectra which are identical in every way with those of model compounds. The PES data, corrected in the usual way for inductive effects (+0.03 eV), confirm the absence of groundstate interaction as well. One is tempted, on this basis, to characterize ESO as a molecule in which photoisomerization is exclusively the consequence of triplet localization a t the double bond,65 and the 90’ relationship in this molecule to be one which virtually eliminates ground and singlet state interactions. Conclusions The photochemistry of these cyclic nonconjugated aryl olefins varies from the complex (“diverted di-n-methane”) rearrangement of MBN to the simple E G Z isomerizations of IBN, EMI, and ESO. The relative participation of singlet and triplet states in the photochemistry has been deduced from xenon perturbation data, and compared with the degree of orbital interactions deduced from photoelectron, absorption, and emission spectroscopy. The singlet interactions strongly depend on geometry, with MBN and its angular atomic orbital relationship producing extensive interplay (cf. 11) while ESO and its orthogonal arrangement yield minimal spectral distortion. The triplets are less sensitive to geometry, there being no phosphorescence and significant triplet derived olefin isomerization in all cases. An important feature ofboth M B N and E M I is the extensice nonradiative decay f r o m singlet as well as triplet states; this decay is obviously caused by the double bond but is not due to its role as a “free rotor”. This observation may well prove to be general for nonconjugated n systems (including di-n-methane molecules) and, in fact, radiationless decay isf a s t becoming one of the most striking consequences of intramolecular bichromophoric interactions ,2,3d,g.6.45a Experimental SectionIb Chemicals. Spectral quality solvents were used in all photochemical experiments. n-Pentane, cyclopentane, cyclohexane, and n-hexane were Burdick and Jackson “Distilled i n Glass’’ and used as received. Spectroquality ethanol was prepared by distilling Commercial Solvents Gold Label ethanol from magnesium ethoxide under a nitrogen atmosphere, with the middle cut retained. Diethyl ether and tetra-

/

100:23

/ Nocember 8 , 1978

hydrofuran (TH F) ( M C B analytical reagent grade) were distilled under nitrogen from sodium benzophenone ketyl; benzene, N , N dimethylformamide (DMF), and dimethyl sulfoxide (Me2SO) (MCB analytical reagent grade) were distilled from calcium hydride prior to use. ti-um- I-Phenyl-2-butene (ca. 97% trans, Aldrich) was purified by column chromatography on silver nitrate impregnated silica gel using hexane as eluent or by preparative VPC on column L at 105 OC (60 mL He/min). c,is-2-Heptene ( K and K ; 95.5% cis, 4.0% trans. 0.5% unknown) was purified by VPC on column J at 50 “ C ( I 50 mL He/min) followed by column K at 45 “ C (100 mL He/min); this procedure provides olefin which is a t least 99.9% cis-2-heptene (colu m n I). Instrumentation. Infrared spectra were taken on a Perkin-Elmer 137 or 221 spectrophotometer; ultraviolet spectra were obtained on a Cary 1 5 or Beckman DUR with Gilford 222-A photometer; N M R spectra were taken on a Varian A-60A, Varian XL-100, or PerkinElmer R-32 spectrometer with MeJSi as a reference (proton) or on a Varian CFT-20 spectrometer (carbon). Mass spectra were recorded by the Purdue Chemistry Department Mass Spectroscopy Center on a Hitachi RMU-6A or CE-I I O spectrometer. Emission spectra were recorded on a custom-made right-angle spectrofluorimeter66 using a low-pressure mercury lamp filtered by a 2537-A interference filter. All spectra are fully corrected. Singlet lifetime measurements were made on a modified T R W nanosecond fluorimeter (Model 31A)6’ using a deuterium source and an RCA 1 P28 photomultiplier. Preparative VPC was done on Varian Models A-90-P, A-700, or 90-P chromatographs, analytical work was on Varian 1200 or I400 units using a Hewlett-Packard Model 3380-A digital recording integrator. Columns were as follows: A, stainless steel (ss) 5 ft X 0.125 in. 10% SE-30; B, ss 150 ft X 0.010 in. Golay column with UCON LB-550X: C, aluminum (AI) 5 ft X 0.25 in. 10% Carbowax 20M; D, AI I O ft X 0.375 in. 30% Carbowax 20M; E, AI 7 ft X 0.375 in. 20% Carbowax 20M; F, ss 5 ft X 0.125 in. 10% Carbowax 20M; G , AI I O ft X 0.25 in. 10% Carbowax 20M; H, AI 10 ft X 0.25 i n . 10% SE-30. I. ss 20 ft X 0.125 in. 4% AgBF4-IO% U C O S 2000 in series with a s s 20 ft X 0.1 25 in. 20% /3,P’-oxydiproprionitrile;J , AI I5 ft X 0.375 in. 20% Carbowax 20M; K, ss I2 ft X 0.1 25 in. 10% AgBF4-20% Carbowax 20M; L. AI I O ft X 0.25 i n . 10% AgBF4-20% Carbowax 20M; M, copper 20 ft X 0.25 in. 30% SF-96. Column supports were A, I00/120 A W - D M C S Chromosorb W ; C , G, H, J , 40/60 AW-DMCS Gas Pack W; D, K, L, M, 60/80 AW-DMCS Chromosorb W; E and back half of I , 60/80 AW Chromosorb P; F, 100/120 AW-DMCS Chromosorb W ; front half of I, 100/120 AW-DMCS Gas Pack W. Analyses. Quantum efficiencies were by uranyl oxalate actinometry or by reference to the E to Z isomerization of 1 -phenyl-2-butene7 using column K at 105 “ C (25 mL N;z/min); all data werecorrected for back reaction.68 The equation used for this correction must be modified when photoproduct isomer is present prior to photolysis; the modified form69 is

p’ = a - Bo In

- z ( ) ff

Bo

where $’ is the “true” isomerization or conversion of the compound,

00 is the fraction of isomer before photolysis, and N is the fraction of isomer under observation, at the photostationary state. Analyses of cis- and trans-2-heptene were made on column 1 at ambient temperature (25 mL N>/min). Photochemical Apparatus. Most preparative and quantum yield work employed a Hanovia low-pressure mercury resonance lam (Model 688A-45) made of Vycor and emitting principally 2 5 3 7 - 1 light. Quantitative data were obtained using a rotating turntable with quartz tubes fitted with graded seals. Deoxygenation was via multiple freeze-pump-thaw cycles at less than 5 X mm. Some early work employed Vycor tubes with argon degassing. 2-Methylenebenznorbornene (MBN). MBN was synthesized by the Wittig reaction on benznorbornen-2-one. It was isolated in 72% yield, bp 92-93 “ C (4.5 mm), and had spectral properties identical with those of a known sample.7o 2-Methylene-dl-benznorbornene. The same procedure was used as for MBN except that methyl-d3-triphenylphosphonium bromide (Aldrich) was employed. The dideuterio-MBN was isolated in 65% yield (bp 92-93 “ C (4.5 mm)): N M R (CDC13) 6 7.09 (symmetrical m, 4 H), 3.66 (m, 1 H), 3.38 (m, 1 H), 2.50 (doublet ofdoublets, 1 H), 2.07- I .58 (m, 3 H) (plus residual signals at 5.08 and 4.68 d integrating to 0.09 H); IR (neat) 6.13 p ; mass spectrum (70 eV) m/e 158 (M+).

Morrison et al.

/ Photochemistry of 2-Methylenebenznorbornene

( E ) - and (2)-lsobutylidenebenznorbornene (IBN). A mixture of benznorbornen-2-one (3.0 g, 19.0 mmol), triphenylisobutylphosphonium bromide (10.0 g, 25.0 mmol), and potassium tert-butoxide (2.8 g, 25.0 mmol) in 20 mL of dry tetrahydrofuran was refluxed for 36 h under nitrogen. The mixture was cooled to room temperature, poured into 100 mL of water, and hexane extracted (2 X 50 mL). The extracts were water washed, dried over magnesium sulfate, and concentrated in vacuo. The resultant mixture of triphenylphosphine oxide and hydrocarbon product was placed on a 3.0 X I O cm column packed with aluminum oxide and eluted with hexane. Removal of the hexane in vacuo gave 3.3 g (86.5%) of a colorless liquid. This material consisted of a mixture of ca. 90% ( Z ) -and 10% (E)-2-isobutylidenebenznorbornene. The isomers were separated at 180 OC on VPC column D (60 mL He/min). The E isomer eluted first with essentially base line separation between the isomers. Both isomers were prepared in greater than 99.5% isomeric purity as observed on analytical column B at 145 "C. Z-IBN: IR (neat) 3.32, 3.43, 5.92, 6.85 p ; N M R (CDC13) 6 7.10 (m, 4 H, aromatic), 4.90 (bd, I H , J = 9.2 Hz, vinyl), 3.92 (bs, 1 H , bridgehead), 3.35 (bs, 1 H, bridgehead), 2.90-2.30 (m, 2 H, 3 exo and methine), 1.95-1.60 (m, 3 H , 3 endo and bridge), 0.95 (d, 3 H, J = 7.0 Hz, CH3), 0.87 (d, 3 H, J = 7.0 Hz, CH3); high-resolution mass spectrum nz/e 198.140 (calcd for C15H18, m/e 198.1408). Anal. (Cl5H18) C , H . E-IBN: 1R (neat) 3.32, 3.43, 5.92,6.85 p ; N M R (CDCI3) 6 7.09 (m, 4 H , aromatic), 5.29 (bd, 1 H , J = 9.2 Hz,vinyl), 3.55 (bs, 1 H , bridgehead), 3.41 (bs, 1 H, bridgehead), 2.58-2.05 (ni, 2 H, 3 exoand methine), 2.00-1.64 (m, 3 H, bridge and 3 endo), 0.93 (d, 3 H, J = 7.0 Hz, CH3), 0.77 (d, 3 H, J = 7.0 Hz, CH3); high-resolution mass spectrum m/e 198.141 (calcd for C15H18, m/e 198.1408). Anal. ( C I S H I XC) , H. Assignment of Configuration to IBK. The Z and E assignments rest on the IH and 13C N M R chemical shifts. ( I ) An examination of molecular models indicates that the methyl groups for the E isomer should be more shielded by the aromatic ring than should those of the Z compound. The two methyl doublets are centered at 6 0.91 and 0.85 for the Z and E assignments as made above. (2) The configuration about a trisubstituted olefin can be assigned by use of the y compression phenomenon, whereby a carbon resonance is shifted upfield when compressed by a y carbon substituent, relative to the isomer without this c o m p r e s ~ i o n(Le., ~ ~ ,R ~ ,~will be upfield in A relative to its position in B). The C I resonance for Z-IBN comes at 6 47.5

>J R2

vs.

>9 R2

A

R3

B

whereas this carbon resonates at 6 52.8 in the E isomer (by comparison, the second bridgehead signal is almost invariant: 6 43.2 and 43.6 for Z and E , respectively). The complete 13C N M R spectrum may be found in the thesis of Fred Palensky. endo-2-Methylbenznorbornene (MeBIV). M B N was hydrogenated using a palladium on calcium carbonate catalyst, methanol/acetic acid (2:l) a s solvent, and a Brown? hydrogenator. Preparative VPC on column M provided a pure sample: N M R (CDC13) 6 7.35-7.0 (m, 4 H, aromatic), 3.30 (m, 1 H, bridgehead), 3.15 (m, 1 H, bridgehead), 2.37 (m. 1 H, exo methine), 2.08 (broad doublet of doublets, 1 H, exo methylene H), 1.86 (doublet of quartets, 1 H, endo bridge H), 1.68 (doublet of triplets, 1 H, exo bridge H), 0.54 (d, 3 H , endo methyl), -0.54 (m, 1 H, remaining endo methylene H ) . Anal. ( C I ~ H 1 4C) , H . Photolysis of MBN. MBN (31 2 mg, 2 mmol) was dissolved in 200 ni1. of hexane, saturated for I h with argon, and irradiated with the low-pressure mercury lamp for 2 h, during which time about 20% of the starting material was consumed. The solvent was removed on a rotary evaporator and the oily residue molecularly distilled (60-1 I O OC ( I .5 mm)). Preparative VPC on column M provided pure samples of 1, 2, and 3. (Prolonged irradiations also produce the di-*-methane product of 2, benzo[3.4] tricyclo[3.2.1 .02.7]octene.)73 1-Methylene-3-vinylindan (I). IR (neat): 6. I2 (olefin stretch), IO. I O + 10.93 (monosubstituted olefin), 11.45 p (terminal methylene). The structural assignment follows from the N M R , which shows a fourproton multiplet at 6 7.65-7.05 (aromatic hydrogens); a typical monosubstituted olefin absorption with three eight-line patterns, 61 5.06

7361

Ha = 2.0, J 1 3 = 9.8, J 1 4 = 0.6 Hz), 6 2 5.14 (523 = 17.1, J 2 4 = 0.9 Hz), and 63 5.90 (J34 = 8.2 Hz); 64 3.82 (overlapping doublet of triplets broadened by long-range coupling to H I and H2, J45 = 8.2, J46 = 6.0 Hz); a characteristic geminal group with two quartets of triplets, 65 3.06 (556 = 16.5,557 = J 5 8 = 2.05 Hz) and 66 2.59 (J67 = 1 6 8 = 2.50 Hz); the exocyclic methylene hydrogens as two triplets, broadened due to long-range coupling, 67 5.05,68 5.45 (578 = 0.3 Hz). Mass spectrum (70 eV): nz/e 156 (M+). Anal. (C12H12) C. H . Benzo[6.7]bicyclo[ 3.2. I]octa-2,6-diene (2). This compound was identified by comparison of its spectra with those in the literatureJ4 and those obtained from an authentic sample.75 3-Allyl-lH-indene (3). This compound was identified by comparison of its spectra to those of a sample synthesized by published procedure~.~~ 1-Allyl-lH-indene (4).This compound could not be isolated in pure form and was barely resolved on column B a t 105 "C. I t was formed most efficiently by photolysis of MBN (200 nig) in 25 mL of hexane saturated with xenon (-80 mL), in a Vycor tube irradiated for 4 h in the Rayonet reactor with the 254-nm lamps. Preparative VPC on column E at I59 O C (100 mL He/min) provided 4 as a 5% mixture in 1. Nevertheless, a portion of the allylic methylene and the indene vinylic and benzylic protons of 4 are clearly visible in the 100-MHz spectrum and could be compared to the N M R of an authentic samand to data in the l i t e r a t ~ r e Coinjection .~~ of the authentic sample and the 4/1 mixture on column B verified the assignment. Photolysis of Dideuterio-MBN. Using identical procedures with those described above, the labeled MBN provided samples of the labeled 1,2, 3, and 4 (recovered starLing material show'ed no evidence of scrambling). Assignments of the deuterium were as follows: 1, the triplets due to H7 (6 5.04) and Hg (6 5.45) are absent; 2, the multiplet due to H2 ( 6 6.15) and HI (6 3.27) is absent;78 3, the four-proton multiplet at 6 3.38 (benzylic and allyl methylenes) diminishes to a two-proton singlet lacking the 6-Hz allyl H-vinyl H coupling; 4, the visible allylic methylene hydrogen (6 2.28), which is normally a fiveline multiplet which includes a 14-Hz coupling to its geminal partner, has been reduced to 0.22 H (0.15 H is predicted for the 85% deuterated MBN used in this experiment) and is a 7-Hz triplet (i.e., the 14-Hz geminal coupling has been removed by the presence of a geminal deuterium). The benzylic hydrogen at 6 3.52, normally split by the methylene into a 7-Hz triplet, is collapsed to a broad singlet. 3-Benzyl-2,4-pentanedione. This compound was prepared by addition of benzyl chloride to sodium acetylacetone in dimethyl sulfoxide.79The product was isolated in 40% yield. bp 106- 108.5 OC (OS mm), and had an N M R spectrum identical with that reported.80 2-Acetyl-3-rnethylindene. A solution of 15.0 g (0.079 mol) of 3benzyl-2,4-pentanedione in I50 g of polyphosphoric acid was warmed to 65 OC in a 300-mL three-neck flask equipped with an overhead stirrer and calcium chloride drying tube for I 2 h. Upon cooling to room temperature the viscous, reddish-black solution was diluted with 300 mL of distilled water, extracted with ether (3 X 200 mL), washed with 5% sodium bicarbonate (2 X 200 mL), dried (anhydrous MgS04), concentrated in vacuo, and chromatographed on a 3.3 X 30 cm silica gel column, slurry packed in hexane. The column wa5 eluted with 5% ether-pentane with 100-mL fractions taken. The desired product was obtained from the 8th through 13th fractions with the starting material concentrated in the 4th through 6th fractions. Concentration in vacuo of the product-containing fractions gave 6.45 g (47.5%) of a light yellow oil. VPC analysis on column A (150 OC, 30 mL Nl/min) showed the product to be approximately 99.8% pure: IR (neat) 3.25, 3.40, 6.00, 6.25, 6.35, 6.85 p ; N M R (CDC13) 6 7.60-7.20 (m, 4 H, aromatic), 3.65 (q, 2 H, J = 2.1 Hz, -CH2), 2.50 (t, 3 H. J = 2.1 Hz, vinyl CH3), 2.40 (s, 3 H, CH3); mass spectrum (70 eV) nile 172 (M+). Anal. ( C 1 2 H 1 2 0 C ) , H. (E)-2-Ethylidene-3-rnethylindan(E-EMI). This material was prepared by the general procedure of Hutchins.81 To a solution of 2(Jl2

1362

Joitrnal of the American Chemical Society

acetyl-3-methylindene (7.5 g. 43.6 mmol) and p-toluenesulfonylhydrazine ( I 2.0 g, 64.0 mmol) in I20 mL of a I : I mixture of DMF-sulfolane containing 300 mg ofp-toluenesulfonic acid at 100 OC was added sodium cyanoborohydride (7.6 g, 120.0 mmol). The solution was heated with stirring at 100-105 "C for 4 h under nitrogen. Upon cooling, the reaction mixture was poured into 150 mL of water, ether extracted (2 X 100 mL), washed with water (2 X 50 mL), dried (anhydrous MgSOd), and concentrated in vacuo. The crude product was then passed through a 2.5 X I O cm aluminum oxide (neutral) plug, eluting with hexane to remove residual sulfolane, concentrated in vacuo, and molecularly distilled to give 5.2 g (75.5%) of a colorless liquid. Analysis by VPC (column B, 140 "C) showed this material to be a mixture of E-EM1 (90.0%),Z-EM1 (1.4%), and internal olefin (8.6%). Preparative VPC on column C (140 "C, 100 mL He/min) provided pure E-EMI: IR (neat) 3.25, 3.35, 3.40, 6.30.6.80. 6.90 p ; N M R (CDCI3) 6 7.15 (in, 4 H, aromatic), 5.45 ( m , 1 H, v i n y l ) , 3.70-3.50 ( m , 3 H, benzylic), 1.65 (ddm, 3 H , J = 7.5, 2.0 Hz, vinyl C H j ) , 1.32 (d, 3 H, J = 8.0 Hz, benzylic CH3); mass spectrum (70 eV) I I I / Y 158 (M'). Anal. ( C I z H I 4 )C, H. (Zp2-Ethylidene-3-methylindan (Z-EM).This isomer was prepared by photolyzing 480 mg of E-EM1 (96% E ) in 75 mL of degassed n pentane for 2 h using the low-pressure mercury lamp. Removal of solvent and filtration through neutral aluminum oxide (pentane) gave 390 mg of a light yellow liquid consisting of a 41 :59 ratio of Z and E isomers. Preparative VPC on column D at I55 "C gave >98.5% Z : IR (neat) 3.25, 3.35, 3.40, 3.45, 6.30, 6.80, 6.90 p ; N M R (CDC13) 6 7.18 (s, 4 H, aromatic), 5.47 (m, 1 H, vinyl), 4.10-3.35 (m. 3 H, benzylic), 1.69 (bd. 3 H , J = 7.8 Hz, vinyl CH3), 1.30 (d, 3 H , J = 7.8 Hz, benzylic C H j ) ; mass spectrum (70 eV) nile 158 ( M + ) . Anal. ( C I Z H I C, ~ ) H. Assignment of Configuration to EMI. The E and Z assignment was made on the basis of I3C N M R data, using the y compression phen ~ m e n o n , ~ Ias . ~outlined * above for IBN. The atoms of interest are '2-1 and C-3; the E isomer should have C-I shifted upfield while the Z isomer should have C-3 shifted upfield. In fact C-l comes at d 33.1 for E and 6 35.9 for Z ; C-3 comes at 6 41.7 for E and 6 39.2 for Z. (The complete I3C N M R spectrum may be found in the thesis of Fred Palensky.) 2-Methyleneindan (MI). T o a mixture of methyltriphenylphosphonium iodide (10.0 g, 24.7 mniol) and potassium rert-butoxide (2.8 g, 24.7 mmol) in 100 mL of dry benzene (distilled from CaHz), 2indanone (3.2 g, 24.7 mmol) in 40 mL of dry benzene was added dropwise over I .O h at 25 OC. The resulting mixture a a s stirred at 25 O C for 25 h under nitrogen and poured into 200 mL of water. The benzene layer was separated and rotary evaporated, and the residue passed through an 8.0 X 15.0 cm aluminum oxide plug with hexane. Removal of solvent in vacuo gave 420 mg ( 1 3.4%) of a colorless liquid which was purified by preparative VPC at 145 "C on column E (400 m L He/min): IR (neat) 3.25, 3.47. 3.52, 6.02, 6.80, 6.90 w ; N M R (CDC13) 6 7.20 (s, 4 H, aromatic), 5.10 (p, 2 H, J = 2.3 Hz, vinyl), 3.70 (t, 4 H, J = 2.3 Hz, benzylic CH2). Anal. (CloHlo) C , H . I-Methylindan. This was prepared by hydrogenation of 3-methylindene using 10% Pd/C with ethyl acetate as solvent. VPC analysis (column E, 159 OC, 100 mL He/min) indicated a quantitative conversion, and preparative VPC (column E) provided pure material which had IR and N M R spectra identical with those reported.sz 2-Ethyl-1-methylindan (EMIH). Identical material was made by reduction of EM1 or by reduction of 2-ethyl-l-methylindene, in both cases using 10% Pd/C and dry tetrahydrofuran. Preparative VPC on column G provided EMIH, presumably the cis isomer: IR (neat) 3.35, 6.20, 6.28, 6.78, 6.86 p ; N M R (CDC13) d 7.12 ( 5 , 4 H, aromatic), 3.25-2.60 (m. 3 H, benzylic C H and CH?), 2.30 (m, 1 H, C(H)CH2), 1.42 (m, 2 H, C(Hz)CHj), 1.04 (d, 3 H , J = 7.5 Hz, benzylic CH,), 0.95 ( t , 3 H, J = 8.0 Hz, CH:C(Hj)). Calcd for Cl2HlO:m/e 160.125. Found: wile 160.125. l-Methylene-8,9-benzospiro[3.5]nonane (MSO). A mixture of 8,9-benzospiro[3.5]nonan-l-oneX3(0.50 g, 2.7 mmol), triphenylmethylphosphonium iodide (1.62 g, 4. I mrnol), and potassium tertbutoxide (0.37 g, 3.2 mmol) in 15 mL of dry tetrahydrofuran was refluxed for I8 h under nitrogen. Upon cooling, the reaction mixture was poured into 50 mL of water and extracted with 3 X 50 m L of hexane. The hexane extracts were passed through a 8.0 X 15.0 cm alumina plug followed by an additional 50 mL of hexane, combined, and concentrated in vacuo to give 0.40 g (80.496) of MSO. The material was purified by preparative VP C on column H at 200 "C (60

/

100:23

/ Noceniher 8 , 1978

niL He/min): IR (neat) 3.25, 3.40, 3.50,6.00,6.20,6.35, 6.75,6.92 (CC14) 6 7.30 (m, 1 H , aromatic), 6.95 (m, 3 H , aromatic), 4.78 (t, I H , J = 2.0 Hz, vinyl), 4.62 (t, 1 H, J = 3.0 Hz, vinll), 2.72 (bt, 4 H, J = 8.0 Hz, benzylic CH2 and allylic C H l ) , 2.40- I .60 (m, 6 H, aliphatic CH2); mass spectrum (70 eV) nile 184 (M+). Anal. (C14H16) C, H . (Z)-l-Ethylidene-8,9-benzospiro[3.5]nonane(Z-ESOI. A mixture of 8,9-benzospiro[3.5]nonan-I-one ( 1 .O g, 5.4 mmol), triphcnylctliylphosphonium iodide (8.4 g, 20.0 mmol), and potassium tert-butoxide (2.3 g, 20.5 mmol) in 25 mL of dry tetrahydrofuran was refluxed for I O h under nitrogen. Workup as for MSO provided I .O g (94%) of ESO consisting of 95% Z isomer (as determined by VPC on column E, IO0 mL He/min). The isomers were separated by preparative VPC on column E (170 "C) to provide pure Z-ESO: 1R (neat) 3.40,6.25, 6.78,6.95 p ; N M R (CC14) 6 7.35 (m. 1 H, aromatic), 6.95 ( m , 3 H, aromatic), 5.18 (m, 1 H, vinyl), 2.75 (m, 4 H, allylic CH2 and benzylic CH2). 1.95 (m, 6 H, ring CH2), 1.20 (dt. 3 H. J = 7.0. 2.4 HL, Ctl,); inass spectrum (70 eV) nile 198 (M+). Anal. (C15H,g)C, H. (E)-1-Ethylidene-8,9-benzospiro[3.5]nonane(E-ESO).Portions (20 mL) of a solution of 500 nig of ESO (95% Z ) in 100 mL of hexane were irradiated in 19-mm i.d. Vycor rubes which had been freezepump-thaw degassed and flame sealed. The source was a Rayonet reactor with 254-nm lamps. After 19 h, the solvent was concentrated in vacuo and the residue passed through a 2.5 X 8.0 cm alumina plug to give 485 mg of a 55:45 Z/E ratio of isomers. Preparative VPC on column E (I70 OC, 100 mL He/min) provided pure E-ESO: IR (neat) 3.40, 6.25,6.78,6.95 p ; N M R (CC14) 6 7.25 (m, 1 H , aromatic), 6.95 (m, 3 H, aromatic), 5.00 (m, 1 H, vinyl), 2.70 (m, 4 H, allylic C H 2 and benzylic CH2), 2.5-1.7 (m. 6 H, 4 and 6 ring C H Z ) , 1.55 (bd, 3 H J = 7.0 Hz, CH3); mass spectrum (70 eV) m/e 198 (M+). Anal. (C15H18) C , H . Assignment of Configuration to ESO. The major synthetic isomer was assigned as Z on the basis of the chemical shift and coupling constants observed for the vinyl methyl signal. In this isomer, the methyl doublet is centered at 6 1.20 with 2.4-Hz triplet fine splitting; the E isomer is centered at 6 1.55 with no triplet fine splitting. The upfield signal in the Z isomer is consistent with the methyl's position relative to the benzene ring, whereby appreciable shielding would be anticipated. Likewise, the Z isomer has the methyl and allylic methylene groups in a transoidal relationship, for which a greater longrange coupling would be expecteds4 than for the cisoid arrangement of the E isomer. 6-Methyl-8,9-benzospiran[3.5]nonane (MSOH). A solution of 96 mg of MSO in 20 mL of dry tetrahydrofuran was hydrogenated over 50 mg of 10% palladium on carbon catalyst for 20 h to afford a 95% conversion to MSOH, as analyzed at 130 "C on column F. The product was purified by preparative VPC at 150 " C o n column G (60 mL He/min): IR (neat) 3.38.6.24,6.70,6.90 p; N M R (CDCI,) d 7.40 (m, 1 H, aromatic), 7.00 (m, 3 H, aromatic), 2.70 (m, 2 H, benzylic), 2.30 (m, 3 H , aliphatic), 1.75 (ni, 6 H , aliphatic), 0.60 (d, 3 H, J = 8.5 Hz, CH3). Calcd for C14Hlg: m/e 186.141. Found: 186.141. Xenon Perturbation Studies. The procedures are basically those of Carroll and Quina.I7 For $iSc determinations using 2-heptene, stock solutions of the alkylaromatic were prepared in cyclopentane, to give an absorbance at 254 nm of 2.1-2.8. This solution (50 mL) was transferred to a volumetric flask and cis-2-heptene was added to make a 40-60 m M solution of olefin. Aliquots (4 or 5 mL) were pipetted into quartz photolysis tubes which had previously been matched to &2.0% (using 1 -phenyl-2-butene photoisomerization). Nevertheless, the tubes were matched for each individual study by finding the orientation of each tube in the cell holder, which gave maximum fluorescence emission (relative to that tube whose emission was strongest). After freeze-pump-thaw degassing ( I ) F'/Fo was determined for pairs of tubes with and without heptene; (2) increments of xenon were added to a separate set of tubes, and (a) F was determined by rotating each tube to its maximum emission, (b) f / f o was determined by irradiation of the set followed by rrans-2-heptene analysis by VPC (see analyses above). The procedure for bichromophoric molecules containing internal "product probes" was essentially identical with that described above, but without the need for added heptene and thus without an F'/Fo determination. A more detailed description of the experiment as well as a complete listing of the data for each compound may be found in the thesis of Fred Palensky.ss p; NMR

Morrison et al.

1 Photochemistry of 2-Methylenebenznorbornene

1363

sible involvement of an upper (T2) triplet state. Reference 8b and references cited therein. See also G. Fischer and E. Fischer, Mol. Photochem., 6, 463 (1974); R. Korenstein, K. A. Muszkat, and E. Fischer, J. Photochem., 5, 345 (1976). J. B. Birks, Photochem. Photobiol., 24, 287 (1976); F. A. Carroll, Mol. fhotochem., 8, 113 (1977). For example, xenon enhancement may be due (1) to a prediminantly triplet reaction which is increased by xenon induced intersystem crossing or (2) a predominantly singlet reaction for which @RT is high but h,, is low. We know of no conclusive evidence on this point for bichromophores. There is a claim (J. Tournon and M. A. El-Bayoumi, J. Chem. Phys., 56, 5128 (1972))that benzyl derivatives exhibit a charge transfer induced increase in k,,,, but the basis for the claim is the false assumption of negligible inReferences and Notes ternal conversion. In fact, k,, is significant (34%) for toluene and it is it which increases upon benzylic substitution, an observation which may be (1) (a) Organic Photochemistry. 40. Part 39: M. Palimer and H. Morrison, J. related to the effects on k,, which we are reporting. (Cf. V. M. Berenfeid Chem. Soc., Chem. Commun., 558 (1978). (b) Abstracted from the Doctoral V. A. Krongauz, b v . Akad. Nauk SSR, Ser, Fiz., 32, 1575 (1968).)The and Dissertations of Frank Scuily (1973), Fred Palensky (1977), and Tom Nylund singlet state energies of BN and MBN are comparable (see below) and i f (1978), Purdue University. this is so for the aryl triplet states as well, the known dependence of k,,, (2) Leading references may be found in H. Morrison, V. Tisdale, P. J. Wagner, on the S-T energy gap (among other factors, of course)37implies a comand K-C Liu, J. Am. Chem. SOC..97, 7189 (1975). mon k,,,. We do know that the k,'s for MBN and its dihydro analogue, MeBN, (3) Recent examples include (a) G. L. B. Carlson, F. H. Quina, B. M. Zarnegar, are identical (see below). The intermolecular interaction of olefins with and D. G. Whitten, J. Am. Chem. Soc., 97, 347 (1975);(b) W. Amrein and aryl excited singlet states increases k,, but does not appear to affect k,,, K. Schaffner, Helv. Chim. Acta, 58,397 (1975); (c) M. A. Schexnayder and (L. M. Stephenson and G. S. Hammond. Angew. Chem., lnt. Ed. Engl., 8, P. S.Engei, J. Am. Chem. Soc., 97, 4825 (1975); (d) H. E. Zimmerman and 261 (1969)). L. M. Tolbert, ibid., 97, 5497 (1975); (e) W. Herz, V. iyer, M. G. Nair, and For example, see J. G. Calvert and J. N. Pitts, Jr., "Photochemistry", Wiley, J. Saltiel, ibid, 99, 2704 (1977); (f) T. Hayashi, N. Mataga, T. Inoue, T. New York, N.Y., 1966, p 296. Kaneda, M. Irie, and S. Misumi, lbid., 99, 523 (1977); (9) R. G. Weiss and G. S.Hammond, ;bid., 100, 1172 (1978); (h) L. N. Domeismith, P. D. Moliere, (38) Using &cBN = 0.59, l ~ B N= liMeBN = 13.5 ns. (39) The terms in eq 11 have been defined in the Resuits section. K. N. Houk, R. C. Hahn, and R. P. Johnson, ibid., 100, 2959 (1978). (40) (a) Note that the deviations in these slopes (cf. Results) make it quite (4) T. J. Chuang. R. J. Cox, and K. B. Eisenthal, J. Am. Chem. Soc., 96,6828 Dossible that 1. 2. 3. and the IBN E -* Zreaction share a common s l o w (1974). (0.46) and singietcomponent (56%). (b) Subsequent to the completion'of ( 5 ) See ref 3b for a recent illustrative study. these studies Hammond and Weiss3g have reported that 5(6) W. Ferree, Jr., J. B. Grutzner, and H. Morrison, J. Am. Chem. Soc., 93,5502 (1971); H. Morrison and W. I. Ferree, Jr., Chem. Commun., 268 (1969). ethyiidene[2.2.1] hept-2-ene also undergoes €12 isomerization out of the (7) H. Morrison, J. Pajak, and R. Peiffer. J. Am. Chem. Soc., 93, 3978 (1971); singlet state and in competition with a di-*-methane rearrangement. (41) We recognize that were U;, assumed to be equal to the common value40 H. Morrison and R. Peiffer, ibid., 90, 3428 (1968). of 0.46, four of the five products would be 100% triplet derived. The ob(8) Portions of this research have been reported in preliminary form; cf. (a) vious problem of such an assumption is the low slope for 4 (a slope less F. Scully, J. Grvtzner, and H. Morrison, J. Am. Chem. SOC.,95,5100(1973); than &,is kinetically impossible). A value of &,, = 0.46 requires a k,,, (b) H. Morrison, T. Nylund. and F. Palensky, J. Chem. Soc., Chem. Commun., 4 (1976); (c) H. Morrison and T. Nylund, [bid., 104 (1977). of 1.0 X lo8 s-l; the value for the monochromophoric models such as benznorbornane is ca. 4.3 X l o 7 s-'. (9) H. Morrison and T. Nylund. J. Chem. SOC., Chem. Commun., 785 (42) (a) Even were &, = 0.4641,& a n d 'kd for MBN would still be high (0.46 (1976). and 1.0 X lo8 s-l, respectively). (b) Hammond and Weiss3g suspected that (10) T. Nyiund and H. Morrison, J. Am. Chem. SOC., following paper in this such is the case for 5-ethylidene[2.2.1]hept-2ene but had insufficient data issue. to confirm their intuition. (c) We assume no appreciable difference in the (1 1) A fifth product is benzo[3.4]tricycio[3.2.1.02~7]octene; it is a secondary photophysicai properties of IPN and MBN. This assumption is substantiated photoproduct not observed at low conversions, which is efficiently formed by the identical 4i values for MBN and 2-isopropyiidenebenznorborfrom 2 by a dh-methane rearrangement.1z nene.1° (12) R. C. Hahn and L. J. Rothman, J. Am. Chem. Soc., 91, 2409 (1969). We (43) A brief review may be found in the excellent chapter by R. S. Davidson in are grateful to Professor Hahn for spectra of this compound. "Molecular Association", Vol. 1, R. Foster, Ed.. Academic Press, New York, (13) These values differ slightly from those previously reported;8athey were obtained by a modification of the earlier analytical procedure and are N.Y., 1975, p 302. considered to be more accurate. (44) Note the absence of &c from eq 15 and this conclusion. (14) R. H. Fleming, F. H. Quina, and G. S. Hammond, J. Am. Chem. SOC..96, (45) (a) We have recently obtained xenon perturbation data for 1-phenyl-2-butene isomerization7which indicate that there is appreciable triplet decay without 7738 (1974). olefin isomerization (D. Giacherio, unpublished results). By contrast with (15) Extended irradiation leads to other products which have not been further MBN, there is no evidence for olefin-induced singlet decay. (b) Ketone investigated. We suspect them to be homologues of the MBN products. triplets do form excipiexes with olefins, competitively with energy transfer: (16) Despite the good precision and reproductibility of the quantum efficiency J. A. Gupta and G. S.Hammond, J. Am. Chem. SOC.,98, 1218 (1976), and and photostationary state data, the latter value is a little lower than would references cited therein. We are not aware of similar observations for aryl be predicted by the quantum efficiencies. triplets. (46) E. Heiibronner and H. Martin, Helv. Chim. Acta, 55, 1490 (1972). (47) Data for 14 from E. Haselbach and M. Rossi, Helv. Chim. Acta, 59, 278 (1976). Benznorbornene and norbornene (8.97 eV) were used to calculate (17) F. A. Carroll and F. H. Quina, J. Am. Chem. SOC.,98, 1 (1976). (18) J. B. Birks, "Photophysics of Aromatic Molecules", Wiley, New York, N.Y., the modified basis energies. 1970, p 122. (48) P. Bischof, J. A. Hashmall, E. Heilbronner,and V. Hornung, Helv. Chim. Acta, 52, 1745 (1969). (19) We are grateful to Professor E. Heiibronner for the MBN data. (49) W. H. Inskeep, D. W. Miles, and H. Eyring, J. Am. Chem. Soc., 92, 3866 (20) F. Brogli, E. Giovanni, E. Heiibronner, and R. Schuster, Chem. Ber., 106, 961 (1973). (1970). (21) P. Asmus and M. Kiessinger, Tetrahedron, 30, 2477 (1974). (50) Of the three benzene transitions, typically the E, (185-195 nm, t -50 000) (22) Prolonged irradiations give 2-ethyl-3-methylindene, an apparent product and B1, (200-210 nm, e -7000) are the most intense (BZu,260-270 nm, of a 1,3 hydrogen shift 6 -400). The N-Vl band of 2-methyienenorbornane comes at 195 n m 5 1 (23) We are grateful to Professor T. Fehiner for the 2-methyieneindan, MSO, Since coupling of two transitions increases as (1) their transition moments and MSOH data. increase and (2) their energy difference decrease^,^^,^^ one expects the (24) Surprisingly, we have found very little in the way of information on this aryl primary aryl olefin coupling to occur within the two higher energy aryl transitions olefin in the literature. Its synthesis proved quite troublesome; see Experimental Section for details. (51) L S forster, A Moscowitz, J G Berger, and K Mislow, J Am Chem Soc , (25) For a review, see S.S.Hixson, P. S. Mariano, and H. E. Zimmerman, Chem. 84. 4353 (1962) Rev., 73, 531 (1973). (52) J. N. Murrell, "Theory of the Electronic Spectra of Organic Molecules", (26) Further evidence for the intermediacy of 9 derives from products isolated Wiley, New York, N.Y., 1963. in methanol, which result from 1,3 addition to the cyclopropane r i r ~ g . ~ , ' ~ (53) D. J. Sandman and K . Mislow, J. Am. Chem. Soc.. 91,645 (1969). See also A product analogous to 9 has been isolated from the photolysis of 7G. G. DeAngelis, Tetrahedron, 24, 5469 (1968). methylenedibenzobicyclo[2.2.2]octadiene; cf. S.J. Cristol and G. 0. Mayo, (54) We are examining other benznorbornenyl derivatives to try to observe the J. Org. Chem., 34, 2363 (1969). dual emission necessary to confirm coexistence of two singlets. (27) J. Ipaktschi, Chem. Ber., 105, 1989 (1972). (55) R. Brainard and H. Morrison, J. Am. Chem. Soc., 93, 2685 (1971). (28) H.lwamura and H.Tukada, Chem. Lett., 1045 (1976), and references cited (56) gXyiene. trisubstituted olefins. and indan have triplet energies of 80.4.57 -74,58 and 82.259kcal/mol. respectively. therein. See however, R. 0. Day, V. W. Day, S. J. Fuerniss, J. R. Hohman, and D. M. S. Wheeler, J. Chem. Soc., Chem. Commun., 853 (1976), for (57) S.L. Murov. "Handbook of Photochemistry", Marcel Dekker, New York, further discussions of the triptycene problem. N.Y., 1973. (29) For one possible exception, see S.S.Hixson and J. C. Tausta, Tetrahedron (58) W. R. Flicker, Ph.D. Thesis, California Institute of Technology, 1976. Lett., 2007 (1974). A recent relevant discussion may be found in P. S. (59) M. Grossman, R. Kubela, G. P. Semeluk, and I. Unger, Can. J. Chem., 50, Mariano, D. G. Watson, and E. Bay, Tetrahedron, 33, 11 (1977). 3298 (1972). (30) Acetone is ineffective and reacts with the MBN. (60) Xenon has no effect on the photostationary state of the I-methylindan triplet (31) The successful choice of a sensitizer involves very subtle differences in sensitized isomerization of 2-heptene. energy levels. Toluene induces both singlet and triplet MBN photochem(61) A constant k,,, is again being assumed.36 istry. (62) F. Palensky and H. Morrison, J. Am. Chem. Soc., 99, 3507 (1977). (32) The lack of 4 in the sensitized photoproduct mixture is anomalous; since 163) We do not. however. see evidence for the charae transfer which characits formation is enhanced by xenon (see below) we are exploring the posterized MEN (cf. 11).

Acknowledgments. We are grateful to the U S . Army Research Office, Standard Oil of Ohio, and Phillips Petroleum Co. for support of this research. We are also grateful to Professor Lytle for assistance with the singlet lifetime measurements, Professors Heilbronner and Fehlner for photoelectron spectral data, Professors Grutzner and Jorgensen for helpful discussions of the N M R and PES data, and Professors Hahn and Wilt for samples and spectra.

7364

Journal o f t h e American Chemical Society

(64) Nicely confirmed by the (€/ZlpSs value of 1.00. The extinction coefficients for Zand € a r e 176 and 178 cm-' M-l, respectively. (65) ESO does not phosphoresce, but the significance of this observation is mitigated bt the lack of phosphorescence in the model spiro alkane. (66) R. 9. Brainard. Ph.D. Thesis, Purdue University, 1970. (67) F. E. Lytle, Photochem. Phofobiol., 17, 75 (1973). (68) A. A. Lamola, Ph.D. Thesis, California Institute of Technology, 1965. (69) We are grateful to Dr. W. W. Schloman, Jr., of these laboratories, for developing this equation. The derivation may be found in ref l b (Fred Palensky). (70) G. L. Trittle, Ph.D. Thesis, Purdue University, 1969. (7 1) G. J. Martin, M. L. Martin, and S. Odiot, Org. Magn. Reson., 7 , 2 (1975). (72) J. W. deHaan and L. J. M. van de Ven, Org. Magn. Reson., 5, 147 (1973). 91, 2409 (1969). (73) R. C. Hahn and L. J. Rothman, J. Am. Chem. SOC., (74) J. M. Brown, E. N. Cain, and M. C. Mclvor, J. Chem. SOC.B, 730 (1971). (75) We are grateful to Professor R. C. Hahn of Syracuse University for samples and benzo[3.4]tricy-

/ 100:23 / Nocember 8, 1978

(76) E. Marechal, Bull. SOC.Chim. Fr., 4089 (1969). The NMR spectrum of 3 may be found in ref I b (F.S.). (77) Prepared using a modification of L. Cedheim and F. Eberson, Synthesis, 159 (1973). (78) HI is superimposed on H5, the second bridgehead hydrogen. The missing proton is assigned as H1 because the 5-Hz coupling between H5 and HdaX is retained. (79) The detailed procedure employed was a modification of T. M. Shepherd, Chem. lnd. (iondon),567 (1970). (80) G. Allen and R. A. Dwek, J. Chem. SOC.B, 161 (1966). (81) R. 0.Hutchins. C. A. Milewski, and B. E. Maryanoff, J. Am. Chem. Soc., 95. 3662 - - ~ I -1973'1. , -, (82) F I E . Juge, Jr.. and A . Frey, J. Org. Chem., 35, 1876 (1970). (83) B. M. Trost and M. J. Bogdonowicz, J. Am. Chem. SOC., 93, 3773 (1971). I (84) G. P. Newsoroff and S. Sternhell, Aust. J. Chem., 25, 1669 (1972). (85) Because of space limitations, the reader is referred to the theses of F.S. and F.P for complete quantitative data for all quantum efficiency, photostationary state, quenching, and sensitization experiments. 1

~

Photochemistry of Bichromophoric Molecules. Photochemistry and Photophysics of 2-Methylenebenznorbornene and Related Molecules in Protic Medial Tom Nylund and Harry hlorrison* Contribution from the Department of Chenzistrj, Purdue Uriicersitj,, West Lafayette, Indiana 47907. Heceired April 25, 1978

Abstract: A novel, acid-catalyzed, photochemical anti-Markownikoff addition of methanol and other nucleophiles to 2-methylenebenznorbornene (MBN) is described. The effect of structure on this reaction has been studied using 2-isopropylidenebenznorbornene (IPBN), 6- and 7-methoxy-2-methylenebenznorbornene (MeOMBN), 6- and 7-trifluoromethyl-2-methylenebenznorborncne (CFjMBN), and 1 -methoxymethyl-2-methylene-d~-benznorbornene (MMBN). The chemical and photopli>sicaldata suggest a mechanism involving three sequential intermediates: 27 (the fluorescent singlet state), 28 (a nonfluorescent, intramolecular charge-transfer state), and 29 (a protonated radical cation). Both 28 and 29 provide efficient radiationless decay paths which result in diminished fluorescent yields and lifetimes for the aryl olefins. The roleof acid is especially noteworthy in that it allows for a n otherwise endothermic electron transfer from the double bond to the aromatic ring; a similar role is likely in other acid-assisted photochemical reactions of aromatics.

The accompanying paper la details the photochemistry and photophysics of the title compound (MBN) in hydrocarbon media. During the course of that study we observed that photolysis of MBN in methanol gives rise to a new reaction path, i.e., anti-Markownikoff addition of solvent to the double bond.? The fact that such addition was unprecedented for a nonconjugated olefin, and the potential relationship of this reaction to the greatly increased radiationless decay characteristic of MBIV,la led us to explore in detail the photochemistry and photophysics of MBN in protic media.

Results Photochemistry of 2-Methylenebenznorbornene (MBN)in Methanol. Irradiation of a 0.02 M solution of M B N in methanol with 0.001 N H z S 0 4 at 254 nm with a low-pressure mercury lamp results in the formation of three photoproducts. All were independently synthesized and have been identified as 1 (endo-2-methoxymethylbenznorbornene), 2 (1 -methoxybenzo[6.7]bicyclo[3.2.l]oct-6-ene),and 3 (cis- and Irans- 1-allyl-3-methoxyindan). Quantum efficiencies are = 0.056 f 0.004, 41= 0.027 f 0.001,42 = 0.012 f 0.001,43 = 0.0038 f 0.0002. The positions of deuterium incorporation when methanol-d was used as solvent were ascertained by 'H N M R analysis and are designated by an asterisk in eq I . There was no deuterium incorporation in the recovered MBN. Photochemistry of 2-Methylene-dz-benznorborene. This 0002-7863/78/1500-7364$01 .OO/O

methanol/H+

MBN 1

2

* 3

photolysis was conducted using conditions identical with those described above. The position of the deuterium atoms was determined by IH N M R ; cf. eq 2. Photosensitized Reaction of MBN. Irradiation of a 0.01 M solution of MBN with 2 M p-xylene and 0.001 N H2S04to low conversion gave the following product ratio: 1,5%; 2,71%; 3, 24%. These numbers may be compared with those obtained upon direct photolysis: 1, 63%; 2, 28%; 3,9%. Xenon Perturbation Study of MBN. M B N was irradiated in methanol with xenon and compared to simultaneous pho-

0 1978 American Chemical Society